Writing about religious change

Before 1 return to the question of narrative, the approach 1 take in this book requires that the question of religious change must first be addressed. The most crucial methodological decision in this enterprise involves settling on a concept of religion that allows one to select and group phenomena together in a manner that is as, or nearly as, meaningful to the author and his reader as it would be to ancient agents, that is to say, a concept of religion that is close to the sense that the latter made of their own actions and the objects involved in these actions. Even at this late stage in my argument it is still worth asking ‘why “religion”?’ The answer is because ‘religion’ is a term that helps us to see certain cultural products and their social context from a perspective that is not offered by ‘politics’ or ‘economy’ or ‘culture’. This is not to claim that religion is a cultural (or even natural) universal. Yet it allows me to make comparisons between the here and there, the then and the now, which 1 would not otherwise be able undertake. There are other terms available which might be even more useful, and humanities as studied in European languages need to engage more intensively with concepts from other cultures, not least to gain a better understanding of developments in cultures that did use the concept of religion. However, pragmatically, I do not, for the time being, see any concept of comparable elaboration that might compete with the concept of religion. That does not exclude paying very close attention to the concepts held by those whose religion we analyse. Such conceptualizations, binding together certain practices and distinguishing them from other practices or concepts or institutions, are part and parcel of an approach such as I take, and must not be confined to identifying the concept that comes closest to religion.

In the case of Mediterranean antiquity, there is no ancient concept that could serve as a meaningful guide. The very word religio was not elaborated into a coherent concept until the late republic, and even then it bore only a very slim correspondence to any notion of religion in use today (see Chapter 1 in this volume and Riipke 2010; in contrast, see Barton and Boyarin 2016). For an analysis of historical change, functionalist notions already fix an important variable, namely the uses and purposes of religion, regardless of the meaning attributed by the human agents. Phenomena might change but the function does not. In exchange for paying this price, these notions allow the researcher to discover a large variety of cultural practices as religion. In contrast, culturist definitions of religion treat religion as being fixed in a different way. They take religion to be shared meanings, systems of beliefs, signs, and practices that were established in particular cultural contexts and provide enduring meaning and orientation in the face of change. Understood in this way, religion is a powerful instrument for informing the meaning-making of agents and corresponds rather closely to everyday notions of the traditionalism and the longue duree of religion. Change, slow or abrupt, has to be admitted, but is difficult to explain if it cannot easily be correlated with change of cultural contexts. Typically, the cultural contexts invoked for the stability of the system of rules and symbols are very different from the contexts invoked for change.

With a view to resolving these problems, in Pantheon 1 made my approach explicit right at the beginning and proposed a concept of religion that is based on religious agency, as discussed in Chapter 3 in this volume. This approach has a substantialist basis, as it starts from classifications that are historically observable (‘gods', ‘demons’, ‘ancestors’) but then focuses on agents and the way in which they act in social situations, thus avoiding the need to presuppose ‘religiosity’, ‘piety’, ‘belief’, or the like. Construing religion through this lens emphasizes innovation and appropriation, referring shared meanings and institutions to the background of individual action. Of course, it is knowledge about gods or ways in which to ‘honour’ ancestors that enables individuals to classify experiences as ‘religious’ (Taves 2010, 2016) or allows one to present oneself as ‘pious’ to an audience (Patzelt 2018). But it is the acts of appropriating and reshaping such knowledge that are foregrounded here, rather than the corpus of narratives about a certain god or a set of ritual rules. Of course, such cultural products come into view as religion, but they are not the essence (for a contrasting view, see, for example, Riesebrodt 2010). Rather, they are medial forms, systematizations or elaborations that might or might not be a part of the repertoire of a period's agents and that might, perhaps, be present in material forms, objects that have been sacralized and imbued with memories that might allow for reappropriation in later periods, even after the earlier practices have been discontinued. ‘Lived ancient religion’, as described in Chapter 4 in this volume, was the tool used to further develop this perspective. This approach was taken over from contemporary Religious Studies but developed beyond the study of the ‘everyday’ into a bundle of methods for the study of textual, iconographic, or material, and even architectural, sources to assist in the reconstruction of religion as religion in the making.

Such explicit theorizing comes at a price. If ‘gods’, ‘temple’, ‘altar’, ‘piety’, ‘priest’ are not the real stuff, but are themselves historical notions used for or developed together with phenomena in a wider framework, then this difference needs to be recalled from time to time. This makes for clumsy terminology where speedy narrative might be expected and/or preferred, but it also allows for differentiation and historicization where differences have been buried underneath the easy use of contingent ancient terms as the concepts of our meta-language. Using ‘not unquestionably plausible addressees’ as a broad generic term, rather than gods, allows us to better capture the seamless

Dealing with religious change 141 transition from dead human beings to divine figures, and from early Iron Age burial practices, through ruler cult, to late ancient Christological titles as addressees of religious communication - one of the surprising and more consequential observations of Pantheon that emerged from using this new terminology. The late rise of the concept of ‘votive’ provides another example. Such observations are only possible if the difference between our analytic concepts and the ancient phenomena and conceptualizations is made explicit.